ICAP MARKET DESIGN: SOME BASIC CONSIDERATIONS

Presented by: Mark D. Younger(Based upon slides originally prepared by Roy J.

Shanker)

2

DISCLAIMER The following presentation is a

personal work product and does not necessarily represent the position of any of my clients.

3

THE MARKET DESIGN (ISO/RTO) PROCESS IS A VERY COMPLEX UNDERTAKING: IT REQUIRES THE CONSISTENT BALANCING

OF: HARD PHYSICS-ELECTRICAL ENGINEERING ECONOMIC THEORY STAKEHOLDER INTERESTS REGULATORY/LEGAL/POLITICAL REQUIREMENTS

4

THE PROCESS TENDS TO “COME OFF THE RAILS” IF ANY OF SEVERAL GENERIC MISTAKES ARE MADE: WE OVERLOOK OR OMIT BASIC ELEMENTS

OR FACTS AND HOW THEY INTERACT WE IGNORE THE PHYSICAL REALITY OF

ELECTRIC SYSTEMS WE ATTEMPT TO REACH COMPROMISE IN

AREAS THAT BASICALLY CAN’T BE COMPROMISED

WE IGNORE POLITICAL REALITIES

5

When we “come off the rails” the resulting prices tend to get characterized as Unfair, or “ Unjust and Unreasonable” if they result in increased prices regardless of why they occurred

An important distinction may be Unanticipated versus Unjust

6

Why Capacity Adequacy Markets? Energy Only Markets Work In Theory Basic Assumptions

Competitive market Atomistic buyers and sellers Rational buyers and sellers with knowledge Elastic demand Elastic supply, no barriers to entry

Functional market design Absence of market power

7

Why Capacity Adequacy Markets/Requirements In theory, in a market system, when demand

exceeds supply the market price would be set by the scarcity costs represented by load bidding in its willingness to be interrupted

The system would clear at prices representing scarcity

Suppliers over time would capture scarcity rents sufficient to attract new capital

There would be no need for capacity requirements

This would be the resolution both short and long term

8

We never meet the assumptions of a fully competitive market in electricity—more “workably competitive” Reasonably concentrated buyers/sellers Reasonable short term (long term?)

barriers to entry No short term demand elasticity at

retail/limited real time information Limited long term demand elasticity

9

How to Deal with these Limitations

THE KEY IS HOW WE DESIGN AROUND THESE LIMITATIONS

Two basic errors in approaching this problem

Most of the problems being experienced today with “unjust or unexpected” are a result of failing to recognize basic market failings and address them from day one

The other basic problem is failing to understand the properties of what has been designed

10

These limitations aren’t necessarily a barrier to going to a market based pricing system

They are however a limitation to going to an energy only market system

These limitations provide specific areas of caution in market design You have to get the prices right You have to protect against market power You have to recognize the implications of

inelastic (uninformed) demand and scarcity in the setting of prices

11

Two Basic Problem Areas Focus today on two

related/integrated and very important areas where we are seeing problems based on a failure to both address a basic problem and a failure to understand what has been designed Short term/real time scarcity Long term adequacy

12

Short Term-Real Time Scarcity Historically, in “equilibrium” regulated

electric supply systems were designed to be “short”

The typical design standard was loss of load probability (LOLP) of “once in ten years”

This means we planned systems with a small expectation that demand would exceed supply

13

Short Term-Real Time Scarcity In theory, in a market system, in real

time when demand exceeds supply the market price would be set by the scarcity costs represented by load bidding in its willingness to be interrupted

The system would clear Suppliers over time would capture

scarcity rents sufficient to attract new capital

14

Short Term-Real Time Scarcity However, currently real time demand is

effectively totally inelastic Retail markets don’t see any of the

wholesale price signals There is currently no practical way to

set clearing prices by scarcity There is no “correct” clearing price

between marginal production cost with adequate supply, and interruption with shortage

15

Short Term-Real Time Scarcity To induce new supplies in this

market annual scarcity rents need to equal the levilized cost of installing new peaking capacity

This requires prices of ~$700/MWh for 175 hours/year or ~$42,000/MWh for 2.4 hours/year

16

Short Term-Real Time Scarcity These energy price levels have not

generally been considered acceptable by some market participants

Without these levels of prices marginal capacity resources will not be added

Even with these prices it is a highly unreliable signal to build new capacity because the amount of revenue to cover capacity is too dependent upon random fluctuations in demand.

17

Short Term-Real Time Scarcity Shortage+Inelastic Demand+Known

Occurrence= Suppliers can set any price they want

We know this is unacceptable, it typically leads to price caps

Price caps are “ok” if we establish alternative ways to collect market clearing revenues-E.g. capacity payments or other ancillary service payments

18

Short Term-Real Time Scarcity “WHAT WILL THE MARKET PRICE OF

ENERGY BE WHEN DEMAND EXCEEDS SUPPLY”

Because of market realities the price can be set at whatever we want it to be in this situation

As a result, you have to answer this question first, and design the markets back from the answer

This is the only way to get a consistent and workable market design

19

Short Term-Real Time Scarcity Possible solutions Design the market (IN ADVANCE)

to allow “politically acceptable” energy prices (caps) plus clearing revenues from other sources in an acceptable fashion Ancillary services ICAP, ACAP, operating reserves

20

Short Term-Real Time Scarcity This means at times of scarcity the price will

rise to the cap What does this mean in terms of pricing:

The “spread” between marginal production costs and the energy price cap is a LEGITIMATE revenue for generation—it reflects a proxy for a portion of shortage costs or scarcity rents

Suppliers have to be able to bid to the cap in scarcity situations regardless of marginal production costs

The use of caps has to be coupled in advance with other clearing revenue sources

21

Long Term Adequacy Again, theory suggests that an energy

only market is all that is needed Participants responding to short term

price signals will build new generation and transmission in response to market needs

Scarcity rents in energy prices only, over the business cycle, are sufficient to attract new capital

Get the energy prices right and the rest (e.g. new entry) follows

22

Long Term Adequacy This is the pricing/market structure for other

capital intensive commodities-aluminum, paper etc.

Characteristics of these other capital intensive commodity markets: Undifferentiated commodity Single commodity price Multiple year business cycle as supply and

demand oscillate and adjust-multiple years of scarcity

High Price volatility over cycle, factors or 3-5 variation not unusual

23

Long Term Adequacy Implications for single commodity price-energy

only-electric markets: High price volatility Scarcity for number of years—business cycle Coupled with demand inelasticity no rational “cap”

for price This is exactly the “unexpected” result of a

design like California, the results are consistent with design

We know this is not politically acceptable-no market design rhetoric is going to change this reality

24

Long Term Adequacy Potential Long Term Solution Administratively set reserves; Installed

Capacity Requirements and Market or Equivalent (e.g. mandatory call contracts backed by hard assets)

These designs don’t necessarily encourage reliable performance (DMNC versus Calls)

Typically criticized as inefficient, e.g. a cost to participants to maintain excess/inefficient resources

25

Long Term Adequacy The cost provides insurance against being

unable to serve load The surplus under normal conditions

dampens volatility in the energy markets The Long system never goes through the

business cycle or has a shorter period of scarcity

Directly complements the solution to the short term scarcity problem

Reality is that this is politically correct and acceptable

26

Long Term Adequacy This is a small price to pay to avoid

persistent scarcity for multiple years coupled with regulatory and political intervention

Some market participants recognize this: “ICAP is the price I am willing to pay to have

access to a liquid and competitive energy market” (Manager of large trading operation.)

Sufficient inherent volatility to keep marketers happy

27

Long Term Adequacy This is exactly what California is

proposing too late Mandatory call rights plus RMR RMR payments effectively are capacity

payments The problem is California is starting

during shortage, and the basic premise of insurance is that you want to buy it during a surplus, before you need to collect

28

Long Term Adequacy What does this mean in terms of Market

Design Mandatory capacity markets and

administratively set levels of reserves Need to recognize pricing implications of

paying for insurance in terms of market design Need to couple with short term solution,

recognizing entitlement to payments in excess of marginal production costs during scarcity

29

Additional Observations These types of solutions work best if

you start from a system in surplus Once the system goes “short”, energy

prices won’t “dampen” until you have solved the physical supply issue

With “bad” design, it will be almost impossible to differentiate or partition between the impacts of scarcity and market power

30

IMPLEMENTATION Given the above, what are some of

the basic functions and problems that have to be addressed with the design of an adequacy market.

31

Basic Functions of An Adequacy Market Establish Reliability/Adequacy Target Establish Reserve/Installed Requirement Assign Requirement to Participants Establish Eligibility/Obligation of

Generation to Participate Measure Capacity Provided Match Supply and Demand Penalties for Failure to Meet Requirements

32

Some Alternative Implementations California-none E.g. a boom/bust cycle New York

NYSRC sets LOLP, ISO Reserve Originally DMNC, now unforced capacity Originally 6 month period, now monthly market ISO auction, bi-lateral Locational requirement, no deliverability or

associated property rights, minimum interconnect standard

Monthly pro-rata deficiency Allocated by peak ratio share

33

Alternative Implementations PJM

RAA/ISO sets reserve requirement Unforced capacity for performance Annual requirement but daily obligation, just

shifted to seasonal periods ISO auction, bi-lateral Deliverability Obligation/Rights, and

associated FTR creation Question on treatment of excess injection rights

Daily pro-rata deficiency going to seasonal Peak load ratio share

34

Alternative Implementations GridFlorida (Proposed)

Proposed ICE, energy call options Monthly requirement No reserve requirement value over 100% Auction and bilateral

Other Central Procurement, long term (not

implemented) Call option with reserve level iron in ground

requirements/reserves (not implemented)

35

Important Issues In this context look at four areas

important to capacity market design Generator performance/evaluation Level of deficiency penalties Time step/obligation period Deliverability/property rights

36

Generator Performance-Evaluation There are a range of alternatives

Single measure DMNC Unforced Capacity (PJM) Performance versus price, call options Performance versus LOLP

The closer you come to performance tracking actual system reliability needs the better

Coarse measures like DMNC don’t relate to performance and may actually encourage lower reliability

37

Generator Performance-Evaluation There has to be a way to measure

a generator’s relative contribution to meeting installed capacity requirements

Ideally there would be a direct link between system demand and performance, e.g. energy at times of highest demand

38

Generator Performance-Evaluation The best solutions should emphasize

actual performance, e.g. UCAP or calls or peak related (as long as rules are known at the start of the market) Is more fair in terms compensation Rewards the “good” players/performers Encourages new entry Encourages retirement of “bad”

performers

39

Deficiency Payments Key “enforcement” element in ANY ICAP

system is the structure and level of deficiency payments

This is the engine that drives “proper” behavior-higher is better

The Deficiency Charge must be high enough to induce new entry

Level of Deficiency Charge must recognize that a new generator is not guaranteed the deficiency charge based payment every year

40

Deficiency Payments The “right” level is consistent with a

premium over the costs of new entry- reflecting all the costs and risks associated with new entry

The level of the premium recognizes that capacity prices drop substantially with a small excess of capacity

Right level also has to reflect opportunity costs if there are adjacent markets without similar market structures

41

Deficiency Payments PJM has set this too low New York is more consistent with function

of deficiency payment Current rate in PJM is a vestige of old

environment where participants had mandatory obligations under state jurisdiction The rate was more of a “capacity equalization

payment” than deficiency or penalty rate

42

Deficiency Payments PJM needs to recognize the

“enforcement” aspect of deficiency charges and move to rates reflecting significant premiums over new entry

43

Time Step One of the most important elements and

least understood is the time step-obligation period-performance period of the capacity adequacy markets

There is a trade-off between having a long time-step that is consistent with the lead time for new resources and a shorter time step that provides more flexibility for shifting load (and capacity) obligations

44

Time Step --Continued Period ICAP Obligation must be based upon an

annual ICAP requirement Most of the shortage risk is targeted to brief

part of the year (usually summer). Shorter obligation period can be

accommodated as long as the deficiency penalty is high enough to induce loads to procure capacity during this period

Failure to do so: Will discourage new entry Will encourage migration of capacity out of system

when prices are higher elsewhere

45

Deliverability Goal is to induce generation addition in

the right locations Requirement for new generation will

vary by ISO Ambiguity with respect to these

obligations and rights creates uncertainty for investment

If there is going to be a deliverability requirement, how do you make sure the system allows for competition from new resources